Lyric-based passwords: Enhancing security and recall with AI
In the digital age, text-based passwords remain the cornerstone of user authentication. However, the balance between security and memorability remains a significant challenge. Users often face a dilemma between creating complex passwords that are difficult to remember and simpler ones that are vulne...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KeAi Communications Co., Ltd.
2025-12-01
|
| Series: | Cyber Security and Applications |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2772918425000256 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In the digital age, text-based passwords remain the cornerstone of user authentication. However, the balance between security and memorability remains a significant challenge. Users often face a dilemma between creating complex passwords that are difficult to remember and simpler ones that are vulnerable to attacks.This research introduces a novel approach to password generation by leveraging linguistic patterns from song lyrics and advanced machine learning models. By processing over 5 million lyrics from the AZ Lyrics and Genius datasets, we identify memorable linguistic constructs, such as verb phrases, to create secure and user-friendly passwords. Transformer architectures are employed for password generation, while LSTM-based models assess their security.A web application integrates these features to enhance usability, offering mnemonic aids such as narrative generation and interactive tools for real-time password creation. This system educates users on best practices and simplifies password management through an engaging interface. Comparative studies demonstrate that lyric-based passwords outperform traditional recall and security metrics methods. By balancing usability and robustness, this approach sets a new standard for password management systems and offers a forward-thinking solution to a persistent cybersecurity challenge. |
|---|---|
| ISSN: | 2772-9184 |